# Op Amp - Dual supplies w/ loads

Discussion in 'General Electronics Chat' started by JStitzlein, Feb 7, 2011.

1. ### JStitzlein Thread Starter Member

Dec 6, 2010
53
0
Hi there,

I modeled and simulated the sine to square wave convertor circuit from the worksheet.

I got the output to match what should be expected, with a minor change. I ran into alot of questions while doing this model, mainly with the supplies like:

1. Is this picture only a theoretical model? If R1 voltage is loaded from the supply there would be a shift in the range of the square wave.

2. How do you calculate for load changes / voltage drops when you attach them to the voltage supply. I dont understand my pspice numbers (for current) as they are off by one magnitude. 15V/1K = 15mA not 1.5mA

3. What is the difference between attaching the ends of the split supply with ground vs no ground?

Thanks for your help,

Last edited: Feb 7, 2011
2. ### Adjuster Well-Known Member

Dec 26, 2010
2,147
300
Hello JStitzlein.

1) Any simulation is a theoretical model. I cannot comment on the effects of R1 because it does not seem to be shown on your schematic.

2) The simulation is correctly calculating the resistor voltage drop. The 1.58mA is the current drawn by the op-amp from its positive supply, so the resistor drops only 1.58V.

3)You must always reference the input voltages of an op-amp to its supply voltages. Doing otherwise will result in the input common-mode voltage being undefined. The simulator may or may not model this condition correctly.
The simulator may also refuse to operate and / or declare error messages related to undefined node potentials if the supplies are not ground related.

3. ### Ron H AAC Fanatic!

Apr 14, 2005
7,050
657
A problem I see with the simulation results is that the current through R25 (8.359ma) should come from V25. The total current through V25 should be 8.359mA + 1.580ma = 9.939mA.

EDIT: I just ran this sim on LTspice. Apparently the 741 behavioral models that we are using does not provide load current to the output.

EDIT: I found an LTspice op amp (LT1195) that shows supply currents>1e12!

Last edited: Feb 7, 2011
4. ### JStitzlein Thread Starter Member

Dec 6, 2010
53
0
For #2)
I just assumed that the supply connected to the op amp with a voltage source would function as a standalone source with no draw.

How do you determine how much current the op amp will draw? Plugging in a range of values gives different currents. If i wanted to calculate what resistor I wanted to feed 7Volts to my Noninverting input how would i do that?

thanks,

5. ### JStitzlein Thread Starter Member

Dec 6, 2010
53
0

Loading down the supply would also shift my output range because the supply would be operating at a lower voltage, no?

6. ### SgtWookie Expert

Jul 17, 2007
22,183
1,728
I think you've found the source of global warming.

7. ### Ron H AAC Fanatic!

Apr 14, 2005
7,050
657
LOL! You could be right.
However, I heard that the last time someone turned one of these on was August 14, 2003.

8. ### JStitzlein Thread Starter Member

Dec 6, 2010
53
0
i would turn one on if i had the equipment

9. ### SgtWookie Expert

Jul 17, 2007
22,183
1,728
I recreated your schematic in LTSpice, and it does work - after a fashion.

However, why do you have R26 between V8's positive and ground? That causes a loss in output swing, as you can see in the attached simulation.

File size:
67.1 KB
Views:
19
10. ### Adjuster Well-Known Member

Dec 26, 2010
2,147
300
As Ron H has noted (but I completely overlooked) the Analog Behavioural Model does not give a correct relationship between supply current and output current for this op-amp. If you place a resistance in the supply line in this model, its effect on the maximum output voltage will therefore not be representative of what would happen in practice.

It is also probably not good practice to use a series resistance in the supply line of an op-amp to control its maximum output swing. The variation of supply voltage with signal conditions and loading could lead to unwanted interactions. A constant power supply voltage is preferable.

Are you trying to get equal positive and negative swings for your square wave? If so, there are better ways to achieve this, such as using amplifiers with rail-to-rail output swing, or otherwise limiting the output.

Edit: What has this to do with the voltage at the non-inverting input? Are you trying to vary the supplies to set the mark/space ratio? That would be like putting a screw into a wall by turning the wall. A much better way would be to use a potential divider for the + input, as in the first diagram.

Last edited: Feb 8, 2011